Abstract
Computational communication science is transitioning from an emerging to an established field within communication research, creating a need for proper guidelines and methodological standards. This forum gathers experienced computational communication science scholars to debate the merits and drawbacks of standardization and discuss the tension between innovation, rigor, and inclusion. The assembled perspectives review current standards for data collection, sharing, and documentation, offering best practices for future research. They argue that high standards and inclusive practices can coexist, enhancing creativity and accessibility. By adopting inclusive guidelines, the computational communication science community can welcome diverse scholars, foster innovation, and advance the field collectively.
Since the seminal article of Lazer et al. (2009) calling out the computational turn in the social sciences, communication research has experienced an “unprecedented boost to progress” (van Atteveldt & Peng, 2018, p. 82), and the subfield of computational communication science (CCS) emerged (Hilbert et al., 2019). Three drivers were essential for the rapid development of the field (Waldherr et al., 2021): first, the availability of abundant digital data from social media and other platforms for research; second, quickly growing computing capacities enabling the modeling and analysis of complex data structures; and third, complex research problems such as misinformation, polarization in digital media environments, and rapidly changing models of media production calling for new and interdisciplinary approaches. This led to a notable increase in communication researchers embracing computational methods such as automated text analysis, large-scale network analysis, or computer simulation and including them into their methodological toolkits.
The growth of CCS research has been particularly notable in media and journalism studies in recent years. To name a few, computational approaches have been applied to research the role of emotion in driving engagement with social media content (Vargo & Hopp, 2020), to study gender representation (Andrich et al., 2023; Sjøvaag & Pedersen, 2018) or ethnic stereotypes (Kroon et al., 2021) in news media, to investigate the place and nature of news content (Weber, 2018), and to automatically detect frames on social media (Zhao & Wang, 2023) or bias in news coverage of major events (Waddell, 2019). The intersection of CCS and media research continues to develop as a critical space for scholarship.
We are now reaching a stage where CCS can no longer be called an emerging field. The Computational Methods Division of the International Communication Association (ICA), founded as a new interest group in 2016, grew quickly and reached division status only 4 years later. Today, it is one of the largest divisions of ICA. Globally, many universities have established tenured professorships for computational or digital methods and/or related research centers. The number of CCS publications is constantly growing, not only in dedicated methods journals such as Computational Communication Research, but also in our discipline’s long-standing top journals (Domahidi et al., 2019; Theocharis & Jungherr, 2021; van Atteveldt & Peng, 2018). Finally, the first CCS textbooks have been published introducing the new canon of methods to students and early-career researchers (Haim, 2023; van Atteveldt et al., 2022).
It is fair to say that CCS is at the threshold of becoming an established and institutionalized field and joining mainstream communication research. At this point, however, there is a palpable need for developing proper guidelines and standards for the application of research methods in this area. The discussion about best practices in CCS and how to translate and ensure long-standing social-scientific standards has accompanied CCS researchers since the early beginnings of the field (Mahrt & Scharkow, 2013; Ruths & Pfeffer, 2014). With more and more newcomers entering the field, and the methodological progress developing faster than ever (thinking, for example, about the pace with which deep learning in image classification or large language models evolve), the calls for standardization have become even more pronounced. These include calls for open science practices such as sharing code and data, defining standards for validity and reliability in CCS as well as for how to deal with ethical questions (Geise & Waldherr, 2021).
This forum creates a space to reflect on these developments in this rapidly evolving subfield. We invited experienced CCS scholars to engage in a fundamental debate on the pros and cons of standardization and to discuss the basic tension between innovation, academic rigor, and inclusion: How can CCS generate new advances in the study of journalism and mass communication? What are the best practices for implementing computational methods in journalism and mass communication research? Which standards do we need, and how rigid should they be? And how can we formulate standards in a way that is inclusive rather than exclusionary to diverse researchers and approaches?
The first commentary comes from Shangyuan Wu, who frames our discussion of CCS by focusing on innovative methodological tools that open new opportunities for studying research questions and phenomena at the core of journalism and mass communication research. She also points to obstacles, such as data access, and limitations of computational methods, emphasizing the need to combine them with more in-depth, qualitative analyses.
Highlighting the increasing maturity of the field, Mario Haim calls for more rigor. CCS scholars, he claims, should and can do better regarding open science practices such as sharing data, documenting code, and replicating research. He strongly urges us to nurture our academic skepticism to ensure the quality and robustness of our work and increase recognition of our field in the social sciences.
Mariken van der Velden argues that progress in CCS “should not come at the expense of exacerbating existing inequalities or excluding individuals from the community.” Instead, she claims, the field must adopt inclusive, community-oriented standards which nurture equity and care, create a hospitable environment, and value diverse perspectives.
Finally, Kaiping Chen elaborates on the concept of inclusive standards and how these can be achieved in knowledge partnerships. This entails co-developing standards and guidelines together with colleagues not yet using computational methods, based on their research questions and needs.
The perspectives assembled in this forum examine the current state of standards regarding data collection, data sharing, analysis, and documentation and offer best practices for future research. They make clear how important it is for the further development of CCS as a field to strike a balance between innovation and standardization and between rigor and inclusion. At first sight, these objectives seem contradictory. Raising the bar too high and formulating strict standards might hinder creativity and innovation and discourage newcomers to join the community.
The conversation in this forum shows that this is not a zero-sum game, and that all three aspects can work together, moving CCS forward as a field and as a community. Good practices and high standards regarding the sharing and documentation of data and code can have an inclusive effect by helping other scholars to learn new approaches more easily. Inclusive standards, guidelines, and best practice examples can decrease barriers for new scholars. In turn, by welcoming more diverse scholars into the community and learning from them in knowledge partnerships, we increase opportunities for further innovation of methods and applications. It is on us as scholars, reviewers, journal editors, and publishers to work toward establishing such positive and reinforcing structures.